and-intake

HIGH W007: Insecure credential handling detected in skill instructions.

• Insecure credential handling detected (high risk: 1.00). The prompt explicitly requires preserving commands and environment variables "verbatim" and to record environment details and commands from the raw signal, which forces the LLM to copy any embedded secrets (API keys, tokens, passwords) into its output/backend writes, creating an exfiltration risk.

MEDIUM W011: Third-party content exposure detected (indirect prompt injection risk).

• Third-party content exposure detected (medium risk: 0.65). SKILL.md instructs the agent to preserve and read user-presented external PR content (author, description, URL, check status, and diff-observed behavior) into the work record body, which is outsider-authored free text entering the agent’s LLM context during intake.